Optimization of KDD Cup 99 Dataset for Intrusion Detection Using Hybrid Swarm Intelligence with Random Forest Classifier

Provided by: International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Topic: Security
Format: PDF
Intrusion detection system plays a vital role in system security which operates data in real time that may leads to dimensionality problem. KDD cup 99 which widely used as a benchmark dataset to detect intrusion is analyzed in this paper. The main drawback of the dataset is its redundancy and duplicate records which reduce accuracy and increase False Alarm Rate (FAR), so a data pre-processing is necessary to reduce obscurity and to clean network data. The Data mining algorithms does not overcome the difficulties of dataset problems, this paper proposed a new method of combining swarm intelligence (simplified swarm optimization) with data mining algorithm (random forest) to pre-process the data.

Find By Topic